version 15.1

cd "G:\Experiment Data"
use Experiment4_encoded_data.dta

**Regression of Performance Ratings on SC-IAT Scores (Table 1)
reg PerformanceRating SCIATscore

**The Interactive Effect of Information and Time on Performance Ratings (Table 3)
reg PerformanceRating i.Treat

**Test of the difference between the effect of information and advertising on performance ratings (Table 4)
test 1.Treat = 2.Treat

**IAT Scores Regressed on Information and Advertising (Table 6 )
reg SCIATscore i.Treat

**Descriptive Statistics(please interpret the results in conjunction with the codebook)
**For continuous variable (Table S2)
sum SCIATscore Age

**For categorical variables (Table S2)
tab Gender
tab Region
tab UrbanRural
tab MaritalStatus
tab FertilityStatus
tab Nation
tab Education
tab PoliticalStatus
tab Income

**Balance test across experimental groups (Table S6)
oneway Gender Treat, tabulate
oneway Region Treat, tabulate
oneway UrbanRural Treat, tabulate
oneway Age Treat, tabulate
oneway MaritalStatus Treat, tabulate
oneway FertilityStatus Treat, tabulate
oneway Nation Treat, tabulate
oneway Education Treat, tabulate
oneway PoliticalStatus Treat, tabulate
oneway Income Treat, tabulate
oneway Prior_EMS_use Treat, tabulate
oneway GovernmentTrust Treat, tabulate

**Potential moderators of information effects (Table S8 and Table S12)
// Recoding and Preparing Variables

**Recoding Treat:
//1 if a subject received information or advertising
//0 otherwise
recode Treat (0=0) (1/2=1), generate(Information)

**Recoding Education: 
//1 if 4-year degree or more
//0 if less than 4-year degree
recode Education (0=0) (1/2=1), generate(New_Education)

**Recoding Income: 
//1 if more than ¥100,000 
//0 if less than or equal to ¥100,000
recode Income (0/1=0) (2/3=1), generate(New_Income)

**Recoding GovernmentTrust: 
//1 if trust just about always or most of the time
//0 if trust only some of the time or almost never
recode GovernmentTrust (3/4=0) (1/2=1), generate(Trust)

**Recoding Prior_EMS_use:
//1 if frequent EMS use 
//0 if infrequent use
recode Prior_EMS_use (1/3=0) (4/5=1), generate(NewPriorEMS_use)

**Recoding PoliticalStatus:
//1 if Communist 
//0 if non-Communist
recode PoliticalStatus (0/1=0) (2=1) (3=0), generate(New_PoliticalStatus)

**The Interactive Effect of Information and Potential moderators on Performance Ratings
reg PerformanceRating i.Information##i.New_Education
reg PerformanceRating i.Information##i.New_Income
reg PerformanceRating i.Information##i.Trust
reg PerformanceRating i.Information##i.NewPriorEMS_use
reg PerformanceRating i.Information##i.New_PoliticalStatus
reg PerformanceRating i.Information##i.Region
